Transactions on Energy Systems and Engineering Applications

Adaptive stochastic gradient descent with least angle regression enhanced navigation: intelligent path planning in cluttered environments for autonomous robots

2025-09-15T20:35:57+00:00

In the dynamic realm of Autonomous Mobile Robots (AMRs), ensuring smooth navigation among obstacles is critical, especially as they become increasingly integral to industries such as manufacturing and transportation. Recent advances have introduced several learning models to aid in obstacle avoidance, but many face computational challenges. This research introduces the Adaptive Stochastic Gradient Descent with Least Angle Regression (ASGD-LARS) algorithm, specifically designed to enhance the navigation of AMRs. By carefully considering obstacle orientations, it facilitates quicker decision-making for direction changes. When compared with well-established algorithms like KNN, XG Boost, Naive Bayes, and Logistic Regression, ASGD-LARS consistently performs better in terms of accuracy, computational efficiency, and reliability. This study lays the foundation for the deployment of smarter and more efficient AMRs across diverse industries.

Investigating the impact of diverse PIDs and ESDs in frequency regulation of a wind-diesel hybrid system

2025-03-31T18:31:28+00:00

Microgrids are gaining momentum these days as they can generate the cleaner and affordable electrical energy through renewable energy sources. The renewable energy sources such as wind has enough potential however, its operation is restricted as wind speed highly varies over the period of the day and that is why diesel engine generation is a possible solution to overcome the wind challenges as well as to supply the uninterrupted electrical energy to the customers. This paper presents the design of various PID controllers to match the energy generation with load demand and hence to stabilize the operation of the microgrid for various operating conditions. The performance of the PID controllers is obtained through gains calculation, diverse error values and through dynamic responses of the microgrids obtained through diverse controllers. Further, this paper also shows the impact of diverse energy storage devices (ESD) with PID controllers for the microgrid, and it is observed that PIL-PID with redox flow battery outperform other controllers and ESDs and most suited for various working conditions of the microgrid.

Thermal conductivity determination in Fe78Si9B13/GNP/Epoxy composites by observation of samples and use of ad-hoc software: a new approximation methodology

2025-07-25T19:30:05+00:00

This study investigated the thermal conductivity (k) of composites composed of Fe78Si9B13 microparticles (weight fractions: 10%, 15%, and 25%) and graphene nanoplatelets (GNP) (weight fractions: 0%, 1.0%, and 1.5%) embedded in a transparent epoxy matrix. Nine cylindrical samples (7 mm diameter and 2 mm length) were prepared. Thermal conductivity was determined by measuring the thermal diffusivity using the flash technique and applying the relevant relationship between the two parameters. Because some samples contained pores, the measured diffusivity was corrected for porosity by using a novel method developed by the authors. This method allowed the estimation of the composite percentage porosity based on the Young's modulus (E) of the sample. This correction eliminates the influence of porosity on the calculated diffusivity value, allowing determination of the intrinsic diffusivity of the composite material. Finally, each sample's thermal conductivity was calculated using the diffusivity values. The values of the estimated parameters were compared with those determined by other well-known and established methods, and practically the same results were obtained. These comparative calculations demonstrated the efficiency of the proposed method. The results demonstrate the effectiveness of this method in correcting the effects of porosity on the thermal conductivity measurements in the studied samples.

Evaluating the influence of diethyl ether on performance and emission outputs in KIRLOSKAR TV-I engines fueled with pumpkin seed oil biodiesel

2024-12-13T09:38:18+00:00

This investigation examines the performance and emission characteristics of the KIRLOSKAR TV-I engine utilizing pumpkin seed oil (Cucurbita pepo L) methyl ester blended with 5% diethyl ether (DEE). Various blends containing 10%, 20%, 30%, 40%, and 50% pumpkin seed oil biodiesel were analysed for their chemical and physical properties, including viscosity, density, flash point, cetane number, and oxidation stability, in compliance with ASTM standards. Gas Chromatography-Mass Spectrometry (GC-MS) was employed to determine the fatty acid composition of the biodiesel. Experimental results revealed that the 20% biodiesel blend exhibited superior performance, combustion, and emission characteristics, making it a viable substitute for conventional diesel with minimal engine modifications. Emission analysis of the 20% blend showed a 0.65% reduction in carbon monoxide (CO), a 10.3% decrease in carbon dioxide (CO2), and a 21.1% reduction in nitrogen oxide (NOx) compared to diesel. Notably, blends without additives also demonstrated significant reductions in NOx (25.83%), CO (14.3%), and CO2 (13.8%) emissions, highlighting the environmental benefits of these biodiesel formulations.

Evaluating the generalization capability of MobileNet-based CNNs for temperature prediction in simulated specklegram fiber optic sensors with combined synthetic datasets and data augmentation

2025-07-09T02:09:54+00:00

The development of machine learning algorithms applied to specklegram-based sensors has facilitated the development of novel approaches for measuring several physical variables; however, most of these methods evaluate a single Fiber Specklegram Sensor (FSS) on a limited dataset. This paper assesses the generalization capability of applying these algorithms, in particular, Convolutional Neural Networks (CNN), to the prediction of temperature in simulated FSSs with different characteristics and conditions. This is achieved through the use of multiple combined synthetic datasets and data augmentation. The application of the Finite Element Method (FEM) enables the generation of datasets within the COMSOL Multi-physics software. The datasets are simulated with varying optical parameters, representing different optical fibers. Following the simulation of the datasets and training of selected models by combining them, data augmentation tests are conducted as though they were real fiber optic disturbances. Ultimately, a model is generated incorporating all the combined datasets and data augmentation, demonstrating the capacity of the model for generalization. This showcases the versatility of the computational methodology for evaluating, designing, and adjusting sensors without the need for experimental data. Additionally, it illustrates that a relatively simple model can be adapted to a variety of sensing system scenarios and configurations.